This invention relates generally to vision systems for vehicles and, more particularly, to rearview vision systems which provide the vehicle operator with scenic information in the direction rearward of the vehicle. More particularly, the invention relates to a rearview vision system utilizing image capture devices, such as CMOS imaging arrays and the like.
A long-felt need in the art of vehicle rearview vision systems has been to eliminate exterior rearview mirrors by utilizing image capture devices, such as cameras, in combination with dashboard displays. This would be beneficial because it would reduce wind drag on the vehicle, wind noise and vehicle weight. Furthermore, rearview mirrors protrude a substantial distance from the side of the vehicle, which makes maneuvering in tight spaces more difficult. Image capture devices are capable of positioning in a greater variety of locations on the vehicle, providing more flexibility of vehicle styling. It is further expected that camera systems would greatly reduce the blind spots to the sides and rear of the vehicle common with vehicles equipped with conventional rearview mirror systems. The driver cannot perceive vehicles, objects, or other road users in such blind spots without turning his or her body, which interferes with forward-looking visual activities.
Camera-based rearview vision systems for vehicles have not obtained commercial acceptance. One difficulty with proposed systems has been that they present a large amount of visual information in a manner which is difficult to comprehend. This difficulty arises from many factors. In order to significantly reduce blind spots, multiple image capture devices are typically positioned at various locations on the vehicle. The image of an object behind the equipped vehicle is usually captured by more than one image capture device at a time and displayed in multiple images. This may confuse the driver as to whether more than one object is present. When multiple image capture devices are positioned at different longitudinal locations on the vehicle, objects behind the vehicle are at different distances from the image capture devices. This results in different image sizes for the same object. This effect is especially noticeable for laterally extending images, such as a bridge, highway crosswalk markings, the earth""s horizon, and the like. Such images are at different vertical angles with respect to the image capture devices. This results in different vertical positions on the display causing the elongated image to appear disjointed.
A camera system provides a monocular view of the scene, compared to the binocular, or stereoscopic, view obtained when the scene is viewed through a rearview mirror. This makes the ability to judge distances in a camera system a problem. This effect is most noticeable at distances close to the vehicle where stereoscopic imaging is relied upon extensively by the driver in judging relative locations of objects. Therefore, known camera systems fail to provide to the driver important information where that information is most neededxe2x80x94at small separation distances from surrounding objects.
Another difficulty with camera systems is that, in order to provide a sufficient amount of information, the camera system typically presents the driver with a greatly increased field of view. This improves performance by further reducing blind spots at the side and rear of the vehicle. However, an increased field of view is often obtained by utilizing a wide-angle lens which introduces distortion of the scene and further impairs the ability of the driver to judge distances of objects displayed. The problem with such distortion of the scene is that the driver must concentrate more on the display and take a longer time to interpret and extract the necessary information. This further distracts the driver from the primary visual task of maintaining awareness of vehicles and other objects in the vicinity of the driven vehicle.
The present invention is directed towards enhancing the interpretation of visual information in a rearview vision system by presenting information in a manner which does not require significant concentration of the driver or present distractions to the driver. This is accomplished according to the invention in a rearview vision system having at least two image capture devices positioned on the vehicle and directed rearwardly with respect to the direction of travel of the vehicle. A display is provided for images captured by the image capture devices. The display combines the captured images into an image that would be achieved by a single rearward-looking camera having a view unobstructed by the vehicle. In order to obtain all of the necessary information of activity, not only behind but also along side of the vehicle, the virtual camera should be positioned forward of the driver. The image synthesized from the multiple image capture devices may have a dead space which corresponds with the area occupied by the vehicle. This dead space is useable by the driver""s sense of perspective in judging the location of vehicles behind and along side of the equipped vehicle.
The present invention provides techniques for synthesizing images captured by individual, spatially separated, image capture devices into such ideal image, displayed on the display device. This may be accomplished according to an aspect of the invention by providing at least three image capture devices. At least two of the image capture devices are side image capture devices mounted on opposite sides of the vehicle. At least one of the image capture devices is a center image capture device mounted laterally between the side image capture devices. A display system displays an image synthesized from outputs of the image capture devices. The displayed image includes an image portion from each of the image capture devices. The image portion from the center image capture device is vertically compressed.
It has been discovered that such vertical compression substantially eliminates distortion resulting from the spatial separation between the cameras and can be readily accomplished. In an illustrated embodiment, the image compression is carried out by removing selective ones of the scan lines making up the image portion. A greater number of lines are removed further away from the vertical center of the image.
The compression of the central image portion produces a dead space in the displayed image which may be made to correspond with the area that would be occupied by the vehicle in the view from the single virtual camera. Preferably, perspective lines are included at lateral edges of the dead space which are aligned with the direction of travel of the vehicle and, therefore, appear in parallel with lane markings. This provides visual clues to the driver""s sense of perspective in order to assist in judging distances of objects around the vehicle.
According to another aspect of the invention, image enhancement means are provided for enhancing the displayed image. Such means may be in the form of graphic overlays superimposed on the displayed image. Such graphic overlap may include indicia of the anticipated path of travel of the vehicle which is useful in assisting the driver in guiding the vehicle in reverse directions. Such graphic overlay may include a distance grid indicating distances behind the vehicle of objects juxtaposed with the grid.
These and other objects, advantages, and features of this invention will become apparent by review of the following specification in conjunction with the drawings.